Deep learning has recently become one of the best way to model the complexity of stock movements and to capture non-linear paths, to associate large data, and to reduce noise without an assumption of a pre-specified underlying structure.
Also hyperparameter optimization or HPO (the problem of choosing a set of optimal hyper-parameters for a learning algorithm) has become an increasingly important issue in the field of machine learning for the development of more accurate forecasting models.
This paper explores the potential of HPO in modeling stock returns using a deep neural network by adopting technical indicators and fundamentals examined based on the effect the regularization of dropouts and batch normalization for all input data.
#quantitativefinance #mathematicalfinance #datascience #mathematicaleconomics #deeplearning #machinelearning #computerscience #ai
Read the full article on the publisher's website.
Posted in: Investment & Finance →
Myself
As an Investment Consultant and Specialist, Pompeo Pontone is a Professional Investor with 25 years’ experience in the fields of Investment Management and Quantitative Finance, with advanced expertise in Computer Science and Data Science.